Skip to Main Content

Sales Teams Have an Efficiency Problem

Heading into 2024, sales teams have a lot riding on them. The economy seems to be picking up, and companies are eager for growth in this newly sunny landscape. Unfortunately, the sales representatives who will be responsible for much of this growth find cross-selling and upselling to be challenging and time-consuming. From sifting through thousands of opportunities to dedicating hours reaching out to account owners, many sellers are spending too much time with too little to show for it.

This will be especially problematic in 2024. According to our recent survey of chief revenue officers, cross-sell and upsell are the two most important areas for revenue diversification. Bottom line: Companies can’t afford to let high-intent customers get lost in the mix. They need a true solution, one that provides reliable intent scoring for cross-sell and upsell prospects. That solution: artificial intelligence.

AI Intent Scoring is Ready for Primetime

Only 30% of companies are using AI-powered sales technology in any capacity. Those who know how much time it can save: Hours of manual analysis can be replaced by a few clicks, with results rivalling the best human analysts. While most companies know they should be exploring AI, intent scoring is a practical first use case.

AI has many benefits—but benefit number one is its ability to make sense of an almost infinite number of data points. To build a better intent scoring system, organizations should start by pulling together as many data sources as they can find. Think about the following data sources as you build your AI model:

  • Past deals. To get to good scoring, you should know what “good” looks like. Get your AI intent scoring model off to a good start by connecting the model to your CRM, allowing the model to look at past deals closed and identify trends.
  • Account attributes. What are some qualities of good cross-sell or upsell prospects? Plug in firmographic, demographic, and technographic information into the model to begin scoring.
  • Products. To know what products your prospects might buy, check what they’re signaling interest in. Any demos they’re watching, current products they’ve bought—throw it into the model.
  • Online interactions. Your customers will signal their willingness to buy in part by what they’re browsing on your website, or how they’re interacting with you on your social media platforms. Make sure that’s reflected in your scoring.
  • Paid marketing engagement. Every ad impression tells you something. A good intent scoring model will take this into consideration too.
  • Job postings. What kinds of roles are your customers hiring for? An investment in team members signals a willingness to make other investments too—so pay attention to your customers’ job boards.

The more pieces of data you can bring into your model, the more effective it will be. Use this list as a starting point and throw in anything you think may give you valuable insights.

Keep it Simple

While the model will be complex, its outputs shouldn’t be. We recommend creating a simple 0 to 100 output score, accessible right in your company’s CRM, that shows each prospect’s propensity to buy a specific product. This will allow your reps to easily start from the top prospects and work their way down. From a business perspective, it should also allow you to see how many strong cross-sell or upsell prospects you have, allowing you to make broader staffing decisions, set informed targets, and generally understand your potential.

Of course, no one likes a black box. To help build trust in the model (and to allow your reps to use their own judgement), we recommend that your model share the major positive and negative indicators that helped it reach its decision. When indicators are positive, reps can use them to shape their outreach. When they’re negative, it’s useful information as well: You can see what kind of objections you might face and plan accordingly.

Iterate and Refine

Like any other part of your business processes, your model will be most useful if it’s continually refined. The CRM interface you build should give reps the ability to share feedback on the data they get, making sure you’re creating a maximally useful product for each end user. In the same way, you can use the data you receive to help you refine your strategy, adjusting your ICP based on patterns in intent scores, and then feeding that ICP back into the model. An effective architecture can look something like this:

Why Turn Scoring Over to AI?

Though this process may be a major departure from traditional intent scoring tools, the approach has its benefits. Snowflake has already seen impressive results from an approach like this, saving 27,000 hours in manual efforts and increasing conversion rates from leads to meetings scheduled by 50%. This will have serious implications for revenue even before we get into the new capabilities an organization could build on top of it (hello retention modeling!).

The data shows that, while revenue leaders think they have the right technology to be successful they’re not satisfied with how it’s implemented. Spaulding Ridge has used industry standard tools like Salesforce and Snowflake to build data-driven revenue solutions like this one across industries. If you’re curious about how to increase your sales efficiency through technology, let’s talk.